What I learned from CEOs of Apple and Corning convinced me of 3 things
Sometimes I think there is the data center, and then there is everything else. If your business is not affiliated with the data center, then it is not going to get a premium stock multiple. If it is, the sky is the limit. We know there are moments when the data-center-as-a-theme seems to dissipate. We saw that occur during the DeepSeek affair, when the launch of DeepSeek-R1 was front-page news in January. The group had a March swoon, one that predated “Liberation Day,” the early April announcement of President Donald Trump ‘s so-called reciprocal tariffs. We had a decline in the group when it looked like Amazon had stopped spending as much. There have been multiple bearish forecasts about the impossibility of spending too big to continue. And then, there have been the “Nvidia’s best days are behind them” callouts from the last couple of quarters. Each one has brought a wrath, a hellish scald of selling so vicious that sellers overwhelm buyers in a way that feels as if they are willing to dump their stocks well below where they are selling. It’s almost like they create order imbalances and the amen chorus of analysts who jump off the data train each time. The whole rap seemed to be hanging by a thread. But then, you got a total recharge, like what happened with Oracle last week, and it was incredible to see how stunningly wrong the bears can be. 1. Data center is booming I am a huge data center believer, as you well know, but I had no idea how powerful the theme could be when Oracle got all promotional as only Oracle can. Sure, the doubters surfaced on day two, when we saw that $300 billion of the trumpeted $455 billion remaining performance obligations were OpenAI. RPO is revenue that is contracted but not yet recognized in the financials. Given that Oracle cloud infrastructure revenue was only $10.2 billion as of the fiscal year 2025, which ended May 31, there has been a bit of a leap of faith. But, OpenAI remains a magical name on Wall Street, and you can see where it could come public at a near-trillion dollar market capitalization — yes, the enthusiasm is that great — and then somehow believers just think the money will be there. It’s all astonishing, almost David Blaine-like, because two weeks ago, the bears were circulating a story that Oracle could never raise enough money to build what it wanted to, let alone ten times that, which is pretty much what’s now slated. Now, it will have OpenAI, plus TikTok parent ByteDance, plus the government and other hyperscalers — those big household tech names — that need help as AI expands. It’s difficult to find a word to describe the impact on the complex after the Oracle news. It spreads from Vertiv , which I love again, to Eaton and Broadcom , Arm — which needed recharge and a resuscitation — and the failing Nvidia. Torrent? Firehose? Tsunami. Pick ’em. Money for the group poured from heaven knows where, not the sidelines and not the “dumb” semis, which were incredible, and the viciously out-of-favor Adobe and Salesforce can only provide so many brooks and streams. It did feel like margin buying was at play — and, of course, the ubiquitous two-times option trash played a role, too. By the end of the week, many of the charts were parabolic, led by the unlikely Micron , with AI really confined there to high bandwidth memory (HDM) as the rest of its product retinue lacks real exposure. DRAMs are strong but have a low margin with a low multiple. DRAM stands for dynamic random access memory. It goes without saying, that even as the analysts are, as usual, totally focused on the Federal Reserve — missing the trees for an alleged clear-cut forest of binary thinking — there was so much money to be made with the group that we had to tell members we would take some profits in Broadcom, if not restricted, because the position got too big. A high-quality problem. The Fed? Good for the cyclicals and homebuilders, and taking your eye off the tech/growth prize. Central bankers meet this week and are expected to cut interest rates by 25 basis points — and perhaps, a total of 75 basis points by year-end. One thing is for certain: the AI Data center is much more than a “strong secular trend.” I give you that rambling preamble because of some insights I gleaned in Harrodsburg, Kentucky, last week “hanging out” with Wendell Weeks, CEO of Corning , and Tim Cook, CEO of Apple . It really was hanging out as we did everything but have a beer together. It’s that kind of town by the way. 2. iPhones are in demand First, we must always link AI with the data center, or we miss the wonderment. Cook told me that AI was an “all-in” situation, and he called it perhaps the most profound change in his lifetime. Those who think that Apple isn’t developing or caring about AI are dead wrong. I learned far too much in the Bluegrass state to believe otherwise. If the Apple bears or those lukewarm on the stock were to have a field trip the way I had they would know their tepid stance is going to bite them in the ass. Sorry for the color, but it is the phrase I was eager to write since my visit. Right now, there are these completely pathetic stories about important people departing from Apple AI. To me, they may only be important to those who write them and the non-believers who always have one foot out the door anyway. For the “trade it, don’t own it” majority, the door will soon be closed on said feet. The Club espouses the opposite: “Own it, don’t trade it” when it comes to Apple. To put it bluntly, Cook has the cards as there have been no customers leaving Apple to go to Samsung, at least that Apple can tell. There is a sense that Samsung will lose subs because the trade-in value of previous iPhone models “in the teens — 13, 14, 15” is staying remarkably high. The press, again, must get free iPhones because the trade-in value, plus the carrier discount — especially with T-Mobile — will keep the admittedly meager growth going. It is growth though, and that matters. I do think that Apple is just scratching the surface of many a country with 100 million people — there are more than you think — and 1.5 billion users remain the target audience that all the AI chatbots need, something again Cook didn’t dispute. I would put a “everything’s on the table” postulation on Apple after the recent district court decision in the Alphabet search case basically urges Google to keep paying Apple. I think that you could even see a bidding war for Apple users, coupled with one more service revenue stream. All of this is going to come true. Notice I offered no caveat. It’s going to happen. As I told Cook, “sounds like an ‘own it, don’t trade it situation,'” something that caused the CEO to justifiably grin with pride. When it does happen, whoever gets the business will have more data center demand than it will be able to meet, and the spillover will benefit Oracle. I can’t say that Google’s chatbot Gemini is the shoo-in because I don’t think Apple can say yet. I just say that the price-to-earnings multiple of Apple will look far different a year from now. I see multiple expansion ahead, a lot of it. But, in Harrodsburg — where all the glass for the Apple Watch and the iPhone will soon be made, a remarkable feat that should be noted even by the doubters in the White House — I bury the lead. It’s Corning that’s the data center play. And, it may be the single best way to play the data center for the moment, even as the stock has soared from a “Liberation Day” low of $39 per share to $77, the last 22 points of a parabolic nature. What’s driving it? There is a big campus refresh, which is a carrier spend, that should benefit the lagging Cisco Systems . That was thought to be a huge spur. However, it’s the virtual monopoly on the fiber linkage between Nvidia’s Blackwell and soon-to-be Vera Rubin that Corning owns. It’s monstrous and, as they say, “it’s not in the numbers.” 3. Corning has opportunities Corning’s embedded in a way that makes the Apple glass deal a sideshow. Incredible, that a contract like that could be a sideshow, but the linked quarter here, which thrives on the AI data center, which needs ten times the fiber of the non-data center, was otherworldly. It had the analysts going gaga, but because of the hybrid nature of Corning’s business, which includes a sold-out wafer business for solar and a robust liquid crystal display (LCD) business where it is the world’s leader, it’s covered by a real mishmash of a community. Suffice it to say that you could stack Corning next to Vertiv as worthy rivals and not go wrong with either. All you do is kick yourself when you think of these things because you get caught up in the idea that the data center is so far superior to all other themes that you are wasting your time with maybe a Capital One or an Abbott Laboratories , and you are simply obliterating capital with a Bristol Myers . I do love the aerospace theme, and that has tons of staying power. But you can tell that buying Boeing at a discount can be perilous. Now, I saw so much that titillated me that I had to restrain myself. Right now, the bruiser Nvidia chips are made with miles of copper in them. They burn way too hot and use too much power. Corning’s got enough business with Nvidia now, but if Nvidia can convert from copper to glass inside it could be a possible ten times increase. I know I didn’t think that was possible, but it is a giddy analyst community consensus. You know I got a sense from Cook that one of the most exciting reasons to upgrade your iPhone is a dual front-back camera that can put you in the picture, and, most importantly, a miraculous camera adjustment that can make it so no one is excluded from a selfie. I have no idea how it works, but I know from our selfies it did. Final coda: These two companies are heroes. They are both capable of having huge years. Yet only one, Corning, is being recognized now. Remarkable. (Jim Cramer’s Charitable Trust is long AMZN, NVDA, ETN, AVGO, CRM. AAPL, CSCO, COF, ABT, BMY, BA. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust’s portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
<
